Investigate Python package metadata, dependencies, files, versions, and vulnerability data from published artifacts on PyPI and private registries — without installing packages.

Provides structured, token-efficient output for AI agents alongside rich terminal output for human users.

Note

Transparency note: This project is developed with heavy use of AI coding agents.
Although most code is AI-generated, architecture and design are human-guided and reviewed, and all changes are tested.

Features

• 🔍 Inspect without installing — query metadata, dependencies, and files from published artifacts — no install required.
• 🌳 Dependency analysis — resolve dependency trees, detect conflicts, and compare versions.
• 📄 Read package files — inspect pyproject.toml, LICENSE, and other files directly from a published distribution.
• 🛡️ Vulnerability scanning — check packages against the OSV database for known security vulnerabilities.
• 🤖 Built for AI agents — built-in agent skill for tool discovery and --format agent for structured, token-efficient output.
• 🔒 Private registry support — works with any PEP 503-compatible package index via --index-url.
• ⚡ Persistent caching — avoid redundant network requests with an SQLite-backed local cache.

peeq vs pip / uv

Task	pip / uv	peeq
View package metadata	Install, then pip show / uv pip show	peeq info <pkg>
View dependencies	Download artifact, extract, parse metadata	peeq deps <pkg>
Compare deps across versions	Custom scripts	peeq deps <pkg> --version X --diff Y
Read a file from a package	Download, extract, navigate to file	peeq cat <pkg> pyproject.toml
Check for vulnerabilities	Separate vulnerability scanner	peeq vulns <pkg>
Resolve dependency tree	Trial install in isolated environment	peeq resolve "pkg>=1.0"
Check for conflicts	Custom scripts or trial-and-error installs	peeq conflicts "pkgA" "pkgB"
Trace why a package is needed	Manual dependency graph tracing	peeq why "requests>=2.31" -d urllib3

Installation

Requires Python 3.10+.

# Install globally with uv (recommended)
uv tool install peeq

# Install with pip
pip install peeq

Zero-install with uvx

If you have uv installed, you can also run peeq directly without installing it.
For example:

uvx peeq info requests

Important

uv is the recommended installation method.
Additionally, commands like resolve, conflicts, and why utilize uv and require it to be installed. See the installation docs for details.

Commands

Command	Description
info	Show package metadata with optional sections
versions	List available versions with filtering
deps	Show dependencies with diff and wheel tag support
artifacts	List distribution artifacts (wheels, sdists) for a version
cat	Print a file from inside a package archive
ls	List paths inside a package archive
download	Download a package archive
vulns	Check for known vulnerabilities (OSV)
resolve	Resolve full dependency tree
conflicts	Check if packages can coexist
why	Trace why a package is in the dependency tree
cache	Cache management (info, clear, check, dump)
config	Configuration management (file path)
skill	Show agent skill instructions

Full documentation

Agent Skill

peeq includes a built-in Agent Skill — structured instructions that teach AI agents how to use peeq for package research.

Install

Download SKILL.md and place it in a peeq/ directory within your agent platform's skill directory.

See the skill documentation for integration options.

Tip

Agents can also load peeq's skill by running peeq skill show directly.

Contributing

See CONTRIBUTING.md for development setup, coding standards, and testing guidelines.

License

MIT